Social security reform is one of the most pressing issues in economic and policy debates around the world. The potential gain from replacing a pay-as-you-go system depends on the gains in efficiency and investment returns from privatization. However, as in any product market, the fundamental behavior of consumers and firms determines market efficiency. The project will examine how workers in Mexico's privatized social security system choose investment funds, and the implication of consumer behavior for competition between firms, equilibrium fees and efficiency. Through a unique collaboration with the Mexican social security administration, CONSAR, I have been given unprecedented access to administrative data on every individual retirement account in the system. First, I will use these data to estimate a flexible discrete choice model of the determinants of workers'fund choices (demand) and how consumers of varying demographics are willing to trade-off fund characteristics, such as load fees, balance fees, past returns, brand name, and convenience (e.g., bank branch proximity). I will then use the demand estimates to examine if the high equilibrium fees in the Mexican pension fund market are due to inelastic demand or non-competitive behavior by firms. Second, I will implement a household survey of randomly selected workers to collect information on key demographics such as education, savings behavior, and measures of financial literacy. I will link the survey data back to the administrative data to examine how these demographics influence worker's choices. In addition, the survey will embed field experiments to test how workers'fund choices change with transparent information on fees and brief tutorials on financial concepts (e.g., the importance of compounding). This will allow me to test if low-income workers pay less attention to fees than higher- income workers because they face higher information or decision making costs, and whether a simple policy intervention can lower information costs and decrease the management fees they pay. This project will bring together unique data sets and diverse empirical techniques to gain an integral understanding of the factors that influence how workers choose funds, how fund managers choose management fees, and the implications for the distribution of wealth at retirement in a privatized social security system. PUBLIC HEALTH RELEVANCE: This research project uses data from the Mexican social security system to examine how workers choose investment funds, how fund managers choose management fees, and the implications for the distribution of wealth at retirement in a privatized social security system. The project also implements a survey and a randomized field experiment in order to examine how decision-making costs and financial literacy impact worker's fund choices, and what policy makers can do to increase competition in a privatized market in order to ensure that workers from all backgrounds can have sufficient funds for a healthy retirement.